Anthropic Mythos AI Audit Finds No New Critical Vulnerabilities in Zcash Protocol

According to Cointelegraph, a security audit of Zcash's protocol conducted by Anthropic's Mythos AI model and commissioned by Shielded Labs found no new critical vulnerabilities, according to Zcash founder Zooko Wilcox.

The audit comes after security researcher Taylor Hornby discovered a four-year-old counterfeit forgery vulnerability in the Orchard shielded pool using Claude Opus 4.8 on June 3. Zcash developers paused Orchard transactions that day and completed a fix shortly after. The Zcash Foundation confirmed no evidence exists that the vulnerability was ever exploited, and user privacy remains unaffected.